Biosynthesis of sporothriolides and sporochartines in fungi

Biosynthesis of Sporothriolides and Sporochartines in Fungi

Von der Naturwissenschaftlichen Fakultät der Gottfried Wilhelm Leibniz Universität Hannover

zur Erlangung des Grades

Doktor der Naturwissenschaften (Dr. rer. nat.) genehmigte Dissertation

von

Dongsong Tian, Master (China)

2021

Referent: Prof. Dr. Russell J. Cox Korreferent: Prof. Dr. Marc Stadler Tag der Promotion: 14.09.2021

Abstract

Key words: Alkyl citrates, biosynthesis, sporothriolides, sporochartines, polyene

The project focused on understanding the biosynthesis of sporothriolides, sporochartines and trienylfuranol A. Gene cluster identification, gene knock out, heterologous expression and protein in vitro assays were used during the investigation.

Alkyl citrate biosynthetic gene clusters of the antifungal metabolite sporothriolide 1 were identified from the genomes of the ascomycetes: Hypomontagnella monticulosa MUCL 54604, H. spongiphila CLL 205 and H. submonticulosa DAOMC 242471. A transformation protocol was established, and genes encoding a fatty acid synthase subunit and a citrate synthase were simultaneously knocked out which led to the loss of sporothriolide and sporochartine production.

Heterologous expression of the spo genes in Aspergillus oryzae then led to the production of intermediates and shunts and delineation of a new fungal biosynthetic pathway originating in fatty acid biosynthesis. Finally, a hydrolase was revealed by in vitro studies likely contributing towards self-resistance of the producer organism. In vitro reactions showed that the sporochartines are derived from non-enzymatic Diels-Alder cycloaddition of 1 and trienylfuranol A 2 during the fermentation and extraction process.

Several hrPKS gene clusters were identified as the potential polyene BGC for trienylfuranol A 2 through multiple bioinformatic analysis, however metabolites produced from the PKS in heterologous expression belong to either different polyene type compounds or pyrone derivatives.

Based on these results, a highly unusual epoxidation/decarboxylation mechanism was proposed to be involved during trienylfuranol A 2 biosynthesis, and a new pyrone BGC likely to encode the biosynthesis of a large class of bioactive compounds related to islandic acid 161 was identified.

Zusammenfassung

Schlagwörter: Alkyl citrates, biosynthesis, sporothriolides, sporochartines, polyene

Das Projekt fokusiert darauf, die Biosynthese der pilzlichen Sekundärmetabolite Sporothriolides, Sporochartines und Trienylfuranol A zu verstehen. Methoden zur Identifizierung von Genclustern, Gen-Knockout, heterologe Expression und in vitro Proteinassays wurden während der Untersuchung angewendet.

Die Alkylcitrat-Biosynthesegencluster (spo) des antifungalen Metabolits Sporothriolide 1 wurden in den Genomen der Ascomyceten Hypomontagnella monticulosa MUCL 54604, H. spongiphila CLL 205 and H. submonticulosa DAOMC 242471 identifiziert. Ein Transformationsprotokoll wurde etabliert und Gene, die für eine Fettsäuresynthaseuntereinheit und Citratsynthase kodieren, wurden gleichzeitig ausgeschaltet. Dies führte zu einem Verlust der Sporothriolide- und Sporochartine-Produktion. Heterologe Expression der Gene aus dem spo Gencluster in Aspergillus orzyae führte dann zur Produktion von Intermediaten und Abzweigungsprodukten.

Die Identifizierung dieser Produkte ermöglichte die Beschreibung eines neuen Biosyntheseweges in Pilzen, der von der Fettsäurebiosynthese abgeleitet ist. Darüberhinaus wurde die Funktion eines im Biosynthesegencluster kodierten Enzyms mit Hilfe von in vitro Untersuchungen als Hydrolase aufgeklärt, welche vermutlich zur Selbstresistenz des Produzenten beiträgt. In vitro Reaktionen von Sporothriolide 1 und Trienylfuranol A 2 zeigten, dass Sporochartine mittels nicht- enzymatischer Diels-Alder-Reaktion während des Fermentationsprozesses entstehen.

Im zweiten Teil der Arbeit wurde die Biosynthese von Trienyfuranol A 2 untersucht. Mehrere Kandidaten-Gencluster mit einer reduzierenden Polyketidsynthase als zentrales Gen wurden mit Hilfe von bioinformatischen Analysen identifiziert und mittels heterologer Expression untersucht.

Jedoch besaß keine der produzierten Substanzen strukturelle Ähnlichkeit mit Trienyfuranol A 2.

Stattdessen handelte es sich um andere Arten von Polyen-Strukturen oder Pyron-Derivate.

Unabhängig von diesen Ergebnissen wurde eine Biosyntheseweg für Trienyfuranol A vorgeschlagen, der auf einem ungewöhnlichen Epoxidierungs-Decarboxylierungs-Mechanismus basiert, und ein neues Pyron-BGC wurde identifiziert, das wahrscheinlich die Biosynthese einer großen Klasse bioaktiver Verbindungen ähnlich zu Islandic acid 161 kodiert.

Acknowledgement

I appreciate Prof. Dr. Russell Cox for providing me with the great opportunity to join the fantastic group in Hannover. Thank you for the nice discussions and helpful supervisions of my PhD project. Your professional knowledge, passion, inspiration, patience and encouragement are very impressive.

I would like to thank Prof. Dr. Jamal Ouazzani who is our great collaborator in France, and the talk with you in Hannover was very nice. In addition, I would like to thank Prof. Dr. Marc Stadler from HZI-Braunschweig and Dr. Mark W Sumarah from Canada for the kind gift of fungi strains and useful suggestions. Thank you Dr. Daniel Wibberg and Prof. Dr. Jörn Kalinowski from Bielefeld, your excellent work on the gDNA and RNA sequencing made a good start for my project. Thank you to the great technical team members in OCI and BMWZ, especially to Katja Körner who is always enthusiastic to help others, and to Jörg Fohrer as well as Linn Haase from our great NMR department.

I would like to thank Prof. Dr. Marc Stadler and Prof. Dr. Jakob Franke for being my PhD thesis examiners. And Thanks Henrike and Dr. Kevin Becker for taking their time to read my thesis and giving me helpful suggestions.

Moreover, I would like to thank all the Cox group members. Thank you Dr. Eric Kuhnert, who is always happy to discuss the scientific questions and give me a hand to solve my life puzzles, but next time please use pure Chinese. Thank you Sen, Chongqing, Jin and Yunlong, the together travel, cooking, Mahjong gaming, swimming and basketball playing in the spare time fully charged my battery. Except, I would like to thank Elizabeth, Carsten, Lukas, Lei, Vjaceslavs, Mary, Katharina, Henrike, Maurice, Haili, Jing, Francesco, Steffen and Eman. Also, special thanks to my project collaborator Dr. Tian Cheng for the warm reception in Braunschweig to me and my girlfriend. If there are some names I forget to mention here, please also accept my thanks.

Thanks to the China Scholarship Council to cover my living expense in Germany.

Particularly, I would like to thank my parents Ming-An (明安) and Zi-Mei (自梅), my sister Jin- Ling (金灵), and my old brother Dong-Sheng (东升), who covered my back throughout my study from college to the PhD. In addition, thanks to Chun-Xia (春霞), my girlfriend, you are the backstage hero. It would be impossible to submit the thesis without your supports.

I enjoyed the time in the last four years, I think I will often recall this memorable experience as it means a lot to me.

Abbreviations and Units

ACP acyl carrier protein KI ketosteroid isomerase-like

att site-specific attachment LCMS liquid chromatography mass spectrometry Spectrometry spectrometry

antiSMASH antibiotics & Secondary Metabolite Analysis Shell

MPT malonyl/palmitoyl transferase

bf byssochlamic acid MS mass spectrometry

BGC biosynthetic gene cluster mRNA messenger RNA

BLAST basic local alignment search tool MeOD deuterated methanol

AT acetyltransferase MeOH methanol

bp base pair NAD(P)H nicotinamide adenine dinucleotide (phosphate)

cDNA complementary DNA NEL normalised expression level

C-MeT C-methyltransferase NMR nuclear magnetic resonance

CoA coenzyme A NOESY nuclear overhauser effect spectroscopy

COSY correlation spectroscopy nrPKS non-reducing PKS

carb carbenicillin NRPS non-ribosomal peptide synthetase

cam chloramphenicol ORF open reading frame

CS citrate synthase ory oryzine

CDCl3 deuterated chloroform PCR polymerase chain reaction

DA(ase) Diels Alder(ase) PEG polyethylene glycol

DAD diode array detector PKS polyketide synthase

ddH2O double distilled H20 ppm parts per million

DH dehydratase prPKS partially reducing PKS

DESeq differential expression sequence PUFA polyunsaturated fatty acid

DNA deoxyribonucleic acid PTM polycyclic tetramate macrolactam

EDTA ethylenediaminetetraacetic acid plf piliformic acid

ER enoyl reductase P_amyB amyB promoter

eGFP enhanced green fluorescent protein P450 cytochrome P450

ESI electronspray ionization PEBP phosphatidylethanolamine-binding proteins PPPRPRPprotein EIC extracted ion chromatogram PPTase phosphopantetheinyltransferase

ELSD evaporative light scattering detector PgpdA gpdA promoter FPLC fast protein liquid chromatography RT-PCR reverse transcription PCR

FAS fatty acid synthase RT retention time

FMO FAD-dependent monooxygenase RNA ribonucleic acid

FAD flavin adenine dinucleotide rpm revolutions per minute

gDNA genomic DNA SAM S-adenosyl methionine

HMBC heteronuclear multiple bond correlation SM secondary metabolites

HPLC high performance liquid chromatography SDR short chain dehydrogenase/reductase hrPKS highly reducing polyketide synthase SQHKS squalestatin hexaketide synthase

1H NMR proton NMR SQTKS squalestatin tetraketide synthase

HRMS high resolution mass spectrometry spo sporothriolide HSQC heteronuclear single quantum coherence TIC total ion current

hph hygromycin B resistance TAE tris-acetate-EDTA

IPTG isopropyl-β-D-thiogalactopyranoside TE thiolesterase

ITS internal transcribed spacer TMS tetramethylsilane

Kan kanamycin UV ultra violet

kb kilo base pairs UPLC ultra-performance liquid chromatography

KR ketoacyl reductase WT wild type

KS ketoacyl synthase yFAS yeast FAS

KO knockout YHR yeast homologous recombination

Contents

Abstract………...………I Zusammenfassung………..………II Acknowledgement………..………III Abbreviations and Units……….…………IV Contents……….V

1 Introduction..………1

1.1 Natural Products from Fungi..………...…1

1.1.1 Fungal Fatty Acid Biosynthesis.………...…2

1.1.2 Fungal Polyketide Biosynthesis..………...…4

1.2 The Biosynthesis of Maleidrides and Alkyl Citrates..………...…6

1.2.1 Maleidrides..………...…6

1.2.2 Squalestatin..………...…7

1.2.3 Oryzine..………..………...…8

1.2.4 Hexylcitric Acid Derivatives..………...….9

1.2.5 Piliformic Acid..………...…10

1.3 Techniques Used in Fungal Biosynthesis Investigations……….…11

1.3.1 Isotopic Labelling………...…11

1.3.2 Genome Sequencing………..………...…13

1.3.3 Gene Knockout………...…14

1.3.4 Heterologous expression………..………...…16

1.4 Overall Aims………...…18

2 Biosynthetic Studies of Sporothriolide………...…19

2.1 Introduction………...…19

2.2 Project Aims………...…20

2.3 Results………...…20

2.3.1 Sporothriolide Production from H. monticulosa, H. spongiphila and H. submonticulosa………...…20

2.3.1.1 Producing and Non-producing Conditions………...21

2.3.1.2 Time Course Study of Sporothriolide Production……….….…22

2.3.2 The Identification of Multiforisin H…….………...…23

2.3.3 Acetate Feeding Experiments of Sporothriolide…………...……...…26

2.3.4 Genome and Transcriptome Analysis………...…27

2.3.4.1 Whole Genome Sequencing and antiSMASH Analysis………...27

2.3.4.2 Sporothriolide BGC Analysis………....…28

2.3.4.3 Transcriptomic Analysis……….…...…31

2.3.4.4 MultiGeneBlast………...………...…34

2.3.5 Gene Knockout in H. spongiphila CLL 205………...…35

2.3.5.1 Fungal Transformation of H. spongiphila CLL 205………35

2.3.5.1a Antibiotics Screening and Protoplast Preparation…...…36

2.3.5.1b Transformation with pTH-GS-eGFP……...…...…...………36

2.3.5.2 Knockout of spofasA/spoE……….………...…37

2.3.5.2a Vector Construction for spofasA/spoE Knockout………..…37

2.3.5.2b spofasA/spoE Knockout Workflow………....…….…38

2.3.5.2c spofasA/spoE Knockout Transformant Analysis…….…...39

2.3.5.3 Attempted Knockout of Other Genes……….…40

2.3.5.3a spoG, spoH and spoK Knockout Vector Constructions and Transformation………...…………40

2.3.5.3b s p o G, s p o H a n d s p o K K n o c k o u t T r a n s f o r m a n t Analysis……….……41

2.3.6 Heterologous Expression in A. oryzae NSAR1……...………....…43

2.3.6.1 Expression Vectors and Cloning Strategies………...………43

2.3.6.2 Gene Combinations and Plasmid Constructions………..………..……44

2.3.6.2a Overview of Constructed Plasmids……….….……....…..…44

2.3.6.2b Gene Combinations and A. oryzae NSAR1 Transformation..48

2.3.6.3 Expression of the spo Genes for Early Steps……….…...49

2.3.6.4 Co-expression of Early Step Genes with Later Tailoring Genes…...…50

2.3.6.4a Co-expression of spofasA, spofasB, spoE, spoL and spo...….50

2.3.6.4b Co-expression of spofasA, spofasB, spoE, spoL, spoK and spoG……..…..….………..…51

2.3.6.4c Co-expression of spofasA, spofasB, spoE, spoL, spoK, spoG and spoH………...…56

2.3.6.4d Co-expression of spofasA, spofasB, spoE, spoL, spoK, spoG and spoJ………...…57

2.3.6.4e Co-expression of spofasA, spofasB, spoE, spoL, spoK, spoG, spoH and spoJ………...……...58

2.3.7 In Vitro Activity Assay with SpoG and SpoI………..62

2.3.7.1 Expression, Purification and Activity Assay of SpoG…………...…….62

2.3.7.2 Expression, Purification and Activity Assay of SpoI………...66

2.4 Discussion………...………..68

2.4.1 Production and Labelling Experiment………...…….68

2.4.2 Bioinformatic Analysis………...………..69

2.4.3 Bipartite Knockout………...………..69

2.4.4 Heterologous Expression and Protein In Vitro Assay……….69

2.4.5 Biosynthesis of Other Alkyl Citrates………...……...73

2.5 Conclusion and Prospect………...………76

3 Biosynthetic Studies of Sporochartine..………...………...………78

3.1 Introduction………...………78

3.1.1 Exampl es of Diels -Al der [4+2] Cycl oaddi tion i n Natural Products Biosynthesis………...78

3.1.2 Enzymatic and Non-enzymatic Diels-Alder [4+2] Reactions………80

3.2 Project Aims………...………...81

3.3 Results………...………81

3.3.1 Sporochartine Production……...………...………81

3.3.2 Time Course Study of Sporochartine Production………...…82

3.3.3 Acetate Feeding Experiment………...………...83

3.3.4 In Vitro Spontaneous Diels-Alder Cycloaddition for Sporochartine…………..85

3.4 Discussion………...………..87

3.5 Conclusion and Prospect………...………88

4 Biosynthetic Studies of Trienylfuranol A………...…………..89

4.1 Introduction………...………89

4.1.1 Depudecin…….……….………...…89

4.1.2 Aureonitol…….……….………...…91

4.1.3 Polyenoic Acid….……...………...…92

4.1.4 Bacterial Enediynes……..………...…93

4.2 Project Aims………...………...95

4.3 Results………...…96

4.3.1 Trienylfuranol A Production………...………...96

4.3.2 Acetate Feeding Experiment………...………...98

4.3.3 Mining of Polyketide Polyene hrPKS………..……….101

4.3.3.1 Genome Screening for hrPKS………101

4.3.3.2 hrPKS Domain Analysis and Transcriptomic Analysis………101

4.3.4 Potential Polyene hrPKS BGC Analysis …….………...…………..104

4.3.4.1 ‘Cluster A’ Analysis………...………104

4.3.4.2 ‘Cluster B’ Analysis………...……….106

4.3.5 Gene Knockout and Heterologous Expression of ‘Cluster A’…...………108

4.3.5.1 Gene Knockout of hspPKS1………...……….108

4.3.5.2 Heterologous Expression of hmPKS1 and hmCAR5 in A. oryzae…...109

4.3.5.2a Gene Combinations, Vector Constructions and A. oryzae NSAR1 Transformation……….…..…110

4.3.5.2b Expression of hmPKS1 in A. oryzae………..111

4.3.5.2c Co-expression of hmPKS1 with hmCAR5……….…114

4.3.6 Heterologous Expression of ‘Cluster B’………...…….114

4.3.6.1 Gene Cloning and Vector Constructions………...…….114

4.3.6.2 Expression of hmPKS2 and hmPKS3 in A. oryzae………116

4.4 Discussion………...………..118

4.4.1 Isotopic Labelling Study………...……….118

4.4.2 Genome Mining and Transcriptomic Analysis………...…………119

4.4.3 Biosynthetic Study of The hrPKS BGC ‘Cluster A’.……….………119

4.4.4 Biosynthetic Study of The hrPKS BGC ‘Cluster B’.……….…………122

4.5 Conclusion and Prospect………...………126

5 Overall Conclusion and Outlook………...……….128

6 Experimental………...…131

6.1 Biology………...………...131

6.1.1 DNA and RNA Extraction and Sequencing………..131

6.1.2 Strains and Transformation...………...………132

6.1.2.1 E. coli Transformation………...……….133

6.1.2.2 S. cerevisiae Transformation (Yeast Homologous Recombination)…133 6.1.2.3 A. oryzae Transformation………...…….133

6.1.2.4 Hypomontagnella spongiphila Transformation………134

6.1.3 Primer Sets and Cloning……...………...………..135

6.1.4 Components of Buffers, Solutions, Liquid Medium and Agar………...136

6.2 Chemistry………...………...138

6.2.1 Fermentation and Extraction of Compounds………..138

6.2.1.1 Small Scale………...………138

6.2.1.2 Large Scale………...………139

6.2.2 Analytical LCMS………...……….139

6.2.3 Preparative LCMS………...………..139

6.2.4 HRMS………...………..140

6.2.5 Nuclear Magnetic Resonance (NMR) Analysis……….140

References………...………..142

7 Appendix………...………...149

1 Introduction

1.1 Natural Products from Fungi

Fungi represent an incredibly rich and rather overlooked reservoir of natural products. Research on fungal metabolites dates back to the 1870s, when pigments synthesized in conspicuous mushroom fruiting bodies attracted the attention of organic chemists. The 20th century witnessed the discovery, isolation and chemical characterization of a vast diversity of natural products from fungi.¹ At the same time, the variety of fungal species and the diversity of their habitats, allow the conclusion that fungi continue to be a rich source of new metabolites.

Fungi produce a wide variety of molecules referred to as secondary metabolites (SM), e.g., polyketides, non-ribosomal peptides, terpenes and alkaloids.² While not directly involved in fundamental metabolic processes of growth and energy generation, SM display an array of biological activities that contribute to the survival of the producing organism in an occupied ecological niche, such as mediating communication within one species or between different species defence against competitors, nutrient acquisition, and even symbiotic interactions.³

Figure 1.1 Examples of fungal secondary metabolites.

Not only does the role of SM make them interesting to study, but many SM, including penicillins, statins, and cyclosporins, have been found to have medical applications.⁴ Among fungal natural products, particular interest is given to antimicrobials, due to the reduction in effectiveness of existing antibiotics used to treat bacterial infections, which is seen as a major threat to global

health security.⁵ Penicillins 4 (Figure 1.1) and cephalosporins 5 are β-lactam antibiotics and represent the most widely used antimicrobials in the world: cephalosporins accounting for 28%

and penicillins for 19% of the global market of antibiotics in 2009.⁶ Some antifungals are also produced by fungi, like griseofulvin 6 and echinocandin 7, which have been used as medicines.⁷ Another critical medical application that fungal SM are known for is that of cholesterol-lowering agents, such as lovastatin 8 which is primarily produced by Aspergillus terreus, and mevastatin 9 found in Penicillium citrinum.^8,9 Fungi could also produce some other SM with immunosuppressant activity. A typical example is the non-ribosomal peptide (NRP) cyclosporin 10 produced by Tolypocladium inflatum and widely used to avoid organ rejection in transplant surgery.¹⁰

1.1.1 Fatty Acid Biosynthesis

Fatty acid biosynthesis is a central metabolic pathway that entails the iterative elongation of fatty acid chains through a set of chemical reactions conserved in all kingdoms of life. Despite the fundamentally different FAS architectures of bacteria, plants, fungi and vertebrates, all of them integrate all necessary enzymatic activities together with acyl carrier protein (ACP) domains used for covalent substrate shuttling from one active site to the other.^11,12

The acetyl primer 11 and malonyl elongation substrates 12 are loaded from coenzyme A (CoA) to ACP by acetyltransferase (AT) and malonyl/palmitoyl transferase (MPT) and condensed to acetoacetyl-ACP in a decarboxylative reaction catalyzed by ketoacyl synthase (KS) (Scheme 1.1, steps I ‒ V, intermediates 13 ‒ 16). In three subsequent reaction steps VI ‒ VIII (intermediates 17

‒ 19), the β-carbon groups are processed by ketoacyl reductase (KR), dehydratase (DH), and enoyl reductase (ER), which results in fully saturated acyl-ACP 19 that can serve directly as a primer (Step IX, intermediate 20) for the next condensation reaction. In each reaction cycle, the growing acyl-chain is elongated by two carbon units until it reaches a length of 16 to 18 carbon atoms 21. The fully saturated carbon backbone is then released as a free acid by a thiolesterase (TE) domain (Scheme 1.1, steps X ‒ XII, intermediates 21 ‒ 23).¹¹

Fatty acid synthases (FAS) are classified into two main groups: type I and type II systems. Type I FAS are multi-domain proteins in which catalytic domains are covalently linked, which represent independent biosynthetic factories because they integrate all necessary enzymatic activities in one megasynthase protein.¹¹ Type I FAS are found in fungi and animals. In contrast, type II FAS are a complex of non-covalently linked mono-functional proteins, each enzymatic activity is catalyzed by a unique protein in the dissociated system. Type II FAS are found in bacteria, plants, and parasites. It is usually believed that type I FAS is a more efficient biosynthetic

machine because the enzymatic activities are fused into a single polypeptide and the intermediates do not diffuse from the complex.¹³ However type II systems offer other advantages because the dissociated acyl ACP intermediates can react with many other cellular catalytic systems.

Scheme 1.1 The fatty acid biosynthetic pathway.

Early fungal FAS was encoded by a single gene, but it split into two separate FAS genes (subunits α and β) during fungal evolution (Figure 1.2).¹¹ In the linear domain organization of modern yeast FAS (yFAS), subunit β contains the AT, ER, DH and the majority of the split MPT domain. The α-chain comprises the ACP, KR, KS, PPTase (phosphopantetheinyltransferase) and smaller part of the MPT domain. The two chains assemble into a heterododecameric complex with a barrel shape. Three full sets of enzymatic domains for fatty acid biosynthesis are located in each of the two reaction chambers which are defined by the β-chain (Figure 1.2). Also, three mobile ACP domains are double tethered to the central hub and the reaction chamber walls (Figure 1.2).

Functionally, ACP can be compared to a mobile arm, which supplies substrates to productive sites of an assembly line.¹¹

In the barrel-like architecture of yFAS, the enzymatic domains are architecturally arranged and concentrated to minimize diffusion distances of the consecutive step of a fatty acid synthesis cycle.

Naturally occurring yFAS only efficiently produce a single type of product, saturated fatty acids.¹¹

Figure 1.2 The barrel-like architecture of yeast FAS. A, a central wheel of six α-chains and two domes of three β- chains on each side, which encoded by two genes (FASα and FASβ); B, the barrel (cut-open view) that contains two

reaction chambers with three double tethered ACPs. Pictures are from literature.¹¹

1.1.2 Fungal Polyketide Biosynthesis

Polyketides are assembled from the same building blocks as fatty acids. The chemical reactions and catalytic domains involved in polyketide and fatty acids biosynthesis are closely related to each other, but they do have some differences. For example, the programming of PKS leads to varied chain length as well as the extent of reduction and elimination during the β-processing process. In addition, PKS are capable of employing other unusual starter and extender units for the chain construction and elongation, such as benzoyl-CoA as the starter unit for squalestatin S1 24 biosynthesis.¹⁴ Fungal PKS often have active C-methyltransferase (C-MeT) domain which can methylate the β-carbon before the KR, DH and ER tailoring cycle.¹⁵ What makes the polyketides more diverse is the post modifications after release. On the contrary, FAS always exclusively produce saturated fatty acids.

Fungal PKS known to-date are type I systems. These are divided into 3 main classes: non- reducing PKS (nrPKS) where there are no reductive steps during chain construction (Figure 1.3, e.g. 3-methylorcinaldehyde 25), partially reducing PKS (prPKS) where there is usually only one reduction (through KR domain) during chain extension (Figure 1.3, e.g. 6-methylsalicylic acid 26), and highly reducing PKS (hrPKS) where the level of reduction is varied and subject to a high level of programming control (Figure 1.3, e.g. squalestatin tetraketide 27), they usually possess the full set of modifying C-MeT, KR, DH and ER domains.^11,16

Lovastatin 8 is a polyketide metabolite produced by the fungus Aspergillus terreus and its biosynthesis has been well-studied.^17,18 The iterative fungal hrPKS (LovB) is responsible for the

B

A

core nonaketide assembly. The ER domain in LovB is inactive, but a trans-acting ER (LovC) is functional and interacts with LovB. LovC accepts three intermediates (30, 31, and 33) and catalyses the reduction steps during the biosynthesis of lovastatin 8. Another PKS (LovF) produces a diketide intermediate 39, which is attached to 38 as the final step for lovastatin formation (Scheme 1.2).

Figure 1.3 Fungal polyketides of nrPKS, prPKS and hrPKS.

Scheme 1.2 The biosynthesis of lovastatin 8 as an example of iterative fungal hrPKS, adapted from Kennedy et al., 1999.^17,18 ER domain in brackets means inactivate.

KS AT DH C-MeT (ER) KR ACP

LovB C

LovC ER

1.2 The Biosynthesis of Maleidrides and Alkyl Citrates

1.2.1 Maleidrides

Maleidrides are carbocyclic compounds with one or two maleic anhydride moieties. A well- studied example is byssochlamic acid 46, which was first isolated from the fungus Byssochlamys fulva.^19,20 The Cox group²¹ sequenced the genome of 46 producer B. fulva. Then bioinformatic analysis was performed to identify a likely maleidride BGC, which was validated by knockout (KO) and heterologous expression experiments.

The putative byssochlamic acid BGC contains four core genes encoding: a hrPKS (Bfpks1), a hydrolase (BfL1), a citrate synthase (BfL2) and a methylcitrate dehydratase (BfL3). Then the transcriptomic analysis of the organism under byssochlamic acid producing and non-producing conditions confirmed the maleidride BGC boundary.²¹

Various combinations of gene sets were constructed for heterologous expression experiments in Aspergillus oryzae NSAR1. Co-expression of the hrPKS bfpks1, citrate synthase bfL2, methylcitrate dehydratase bfL3 and hydrolase bfL1 led to the production of both 43 and 44 (Scheme 1.3).

Co-expression of the two ketosteroid isomerase (KI)-like genes (bfL6 and bfL10) with the four core genes (bfpks1, bfL1, bfL2, and bfL3) led to the production of byssochlamic acid 46 and the decarboxylated intermediate 44, as well as the low titre of agnestadride A 49 and the intermediate 43 (Scheme 1.3). It indicated the two KI-like genes catalyze the dimerization of monomers (43 and 45) to form more complicated scaffolds. However, heterologous expression studies showed that single use of either KI is not sufficient to catalyze any dimerization. More interestingly, experiments showed that the two PEBP (phosphatidylethanolamine-binding proteins) enzymes (BfL5 and BfL9) appear to be involved in the dimerization, because higher titres of byssochlamic acid 46 and heptadride 49 were observed when co-expression the PEBP genes with the four core genes and KI-like genes in A. oryzae.

Scheme 1.3 The proposed biosynthesis of maleidrides by Williams et al., 2016.²¹

1.2.2 Squalestatin

Squalestatin S1 24, isolated from Phoma sp., is a potent and selective inhibitor of squalene synthase.²² For the biosynthesis research of squalestatin S1 24 (Scheme 1.4), detailed molecular studies have revealed that a dedicated SQHKS (squalestatin hexaketide synthase) produces a carbon skeleton that is then condensed with oxaloacetate by citrate synthase (CS) to give an early alkyl citrate intermediate 50 that is further oxidatively processed to 51, then 51 is coupled with a tetraketide 52 that is assembled by SQTKS (squalestatin tetraketide synthase), to afford squalestatin S1 24.¹⁴

Byssochlamic acid 46 and squalestatin S1 24 share similar early steps in the biosynthetic pathways, such as the condensation of polyketide and oxaloacetate that catalysed by the key enzyme CS to produce alkyl citrate intermediates (Scheme 1.3 ‒ 1.4). But the following tailoring steps are diverse. For instance, these reactions include dehydration, decarboxylation and

dimerization required for byssochlamic acid 46 formation, and multiple oxidations involved during squalestatin S1 24 biosynthesis.

Scheme 1.4 The proposed biosynthesis of squalestatin S1 24, adapted from Lebe et al., 2019.¹⁴

1.2.3 Oryzine

Oryzines A 53 and B 54 (Figure 1.4) are two secondary metabolites that were isolated from Aspergillus oryzae.²³ They belong to the alkyl citrate type compounds based on the combination of a C8 unit and a C3 unit. Therefore, at least a CS is required to perform the key alkyl citrate backbone construction. Similar early step genes also exist in the byssochlamic acid 46 and squalestatin S1 24 BGC (Scheme 1.3 ‒ 1.4).

Figure 1.4 Structures of oryzine A 53 and oryzine B 54.

A putative oryzine gene cluster was found by CS (BfL2) homology search of A. oryzae genome.

In summary, the gene cluster (Figure 1.5) encodes two fungal FAS subunits (oryfasA and oryfasB);

a citrate synthase (oryE); a methylcitrate dehydratase (oryR); a decarboxylase (oryM); an alpha- ketoglutarate-dependent dioxygenase (oryG) and two lactonases (oryH and oryL). In addition, three transporters (oryC, oryF, oryN), a transcriptional regulator (oryO), a putative dehydrogenase (oryD), an acyl-CoA ligase (oryP) and a P450 (oryQ) can be found in this gene cluster. However,

no experimental evidence could be used to describe the oryzine biosynthetic pathway details before our work.²³

Figure 1.5 Putative oryzine BGC from A. oryzae RIB 40.²³

1.2.4 Hexylcitric Acid Derivatives

Recently, another alkyl citrate gene cluster was identified from the filamentous fungus Aspergillus niger by bioinformatic analysis.²⁴ This BGC encodes FAS subunit alpha (akcA), FAS subunit beta (akcD), citrate synthase (akcB), transcriptional regulator (akcR), 2-methylcitrate dehydratase (akcC) as well as other co-localized functional genes.²⁴ Eleven hexylcitric acids were generated at g∙L^‒1 level through the overexpression of the transcriptional regulator akcR and the hexylaconitic acid decarboxylase gene hadA (outside the alkyl citrate gene cluster), which are defined as artificial production (with genetic manipulation) compared with the previously reported titre (mg∙L^‒1 level) of natural production.

The early steps in the proposed pathway (Scheme 1.5) are similar to the early steps of byssochlamic acid 46 and squalestatin S1 24 biosynthesis (Scheme 1.3 ‒ 1.4). Hexylcitric acid 56 is generated by the condensation of fatty acid unit 55 and oxaloacetate 41. Normally, a dedicated citrate synthase catalyses this reaction, the AkcB takes the role in this pathway. The next step is the dehydration by 2-methylcitrate dehydratase homolog (AkcC) to make 57, then decarboxylation by HadA to produce 58. However, the stereochemical courses of citrate synthase (AkcB) and 2-methylcitrate dehydratase (AkcC) are unknown in the proposed pathway due to the lack of NMR spectra evidence of these intermediates.²⁴

Scheme 1.5 Proposed biosynthetic pathway for the hexylcitric acid and hexylaconitic acid by Palys et al., 2020.²⁴

Transcription regulator Decarboxylase

2-methylcitrate dehydratase Unknown

Transporter SDR FAS

Citrate synthase Dioxygenase

Lactonase Acyl-CoA ligase

P450

1.2.5 Piliformic Acid

Piliformic acid 59 was isolated as a secondary metabolite from several closely related fungi of xylariaceous genera.^25,26 Piliformic acid 59 requires a C8 unit (octanoate chain) and a C3 moiety for the construction of the scaffold.

The octanoate chain could be assembled from a FAS. Alternatively, a dedicated PKS may build a carbon chain for secondary metabolite biosynthesis. Extensive isotopic labelling investigations were carried out by the O’Hagan group.²⁷ Through the observations of the stereochemical location of deuterium in the octanoate chain, the origin of the C8 unit was concluded to be from a FAS.

This can be determined because the stereochemical course of ER in fungal FAS (pro-R labelled) and fungal PKS (pro-S labelled) are opposite.²⁷

However, whether the octanoate is biosynthesised de novo for secondary metabolism, which means from a specific short-chain FAS, or possibly octanoate is a result of β-oxidation of the higher fatty acids which are synthesised by a FAS of primary metabolism was not known.

O’Hagan group designed the labelling experiment by supplementing isotopically labelled [3-¹³C]

decanoate.²⁸ Then, the incorporation patterns of the carbons from the C-1 acetate were observed, which could be well explained that the β-oxidation of the labelled material yields [1-¹³C] acetate and then incorporation de novo for piliformic acid 59 synthesis (Scheme 1.6). In the contrast, the only [1-¹³C] octanoate generated from the β-oxidation of [3-¹³C] decanoate was not observed.

These results nicely proved the C8 unit of piliformic acid is derived from a FAS with a sole function of octanoate production for 59 biosynthesis.

In addition, the C3 unit is indicated to be derived from the citric acid cycle intermediate oxaloacetate by the efficient incorporation of labelled succinate.²⁷ In a word, the manner of natural products biosynthesis study before the genomic age was achieved predominately by utilizing isotopic labelling.

Scheme 1.6 The labelling patterns for piliformic acid 59.^27,28

Based on the knowledge of byssochlamic acid 46 and squalestatin S1 24 biosynthesis, the condensation of the C8 and C3 moieties in piliformic acid 59 is probably also catalysed by a

dedicated CS. However, the dedicated BGC and biosynthetic pathway for piliformic acid were still unknown before our work.

1.3 Techniques Used in Fungal Biosynthesis Investigations

1.3.1 Isotopic Labelling

The long and successful history of isotopic labelling experiments has tremendously changed and deepened our understanding of natural products biosynthesis. Even now, it continues to provide important insights into the biosynthetic pathways of secondary metabolites.^29–31

The ¹³C incorporations are useful for understanding fatty acids and polyketides biosynthesis.

Incorporation of the double-labelled acetate ([1,2-¹³C2]) into the metabolites, a ¹³C-¹³C coupling will be observed in the ¹³C NMR if the labelled acetate bond remains intact throughout the biosynthetic pathway. And the ¹JCC coupling value at C-1 and C-2 are identical (Figure 1.6B).

However, the intensity of the carbon signal will be enhanced when incorporating the single- labelled acetate ([1-¹³C], [2-¹³C]) into the metabolites, as well as when the incorporation of the double-labelled acetate bond break, because of an increase of the ¹³C content at a particular carbon (Figure 1.6C).³²

Moreover, elucidating the origins of hydrogen and oxygen are also vital to understand the mechanism of chemical steps involved in biosynthesis. When incorporation of ¹⁸O is achieved by using a precursor where the ¹⁸O is directly linked to a ¹³C ([1-¹³C, ¹⁸O2] acetate) or by growth in an ¹⁸O2 environment, the presence of the ¹⁸O alpha to ¹³C can be detected in the ¹³C NMR spectrum by an upfield shift of the signals (Figure 1.6D ‒ E). The substitution of a deuterium alpha or beta to ¹³C can also result in a similar upfield shit in ¹³C NMR, but additional with the multiplet from the spin-spin coupling (¹JCD).³²

Figure 1.6 Illustration of feeding labelled precursors and subsequent signal of incorporated atoms in ¹³C NMR spectra, adapted from Simpson, 1987.³²

A good example of isotopic labelling practice is in the study of squalestatin S1 24 biosynthesis.^14,33 First, ¹³C labelled acetate feeding experiment showed the origin of 1 polyketides, from a benzoate-primed hexaketide and a dimethylated tetraketide (Scheme 1.7).³³ Then the incorporation of ¹⁸O ([1-¹³C, ¹⁸O2] acetate or ¹⁸O2 as a precursor) in squalestatin S1 24 revealed that oxygen atoms (red colour) at C-1, C-3, C-5, C-6, C-7 and C-12 are inserted by molecular oxygen not acetate, therefore a few oxidation steps are anticipated to be involved in the biosynthetic pathway. In contrast, the C-24 and C-34 carbonyl oxygen atoms are derived from acetate.³³

Precursor Metabolite

C spectrum A

B

C

D

E

Scheme 1.7 The labelling studies of squalestatin S1 24, adapted from Jones et al., 1992.³³

1.3.2 Genome Sequencing

The development of genome sequencing accelerates the biosynthetic study of microbial natural products. Especially, researchers realize the potential and capability of microorganisms to produce more amounts of SM than previously acknowledged.^2,34

For example, Galagan and co-workers³⁵ reported a high-quality draft sequence (approximately 40-megabases) of the Neurospora crassa genome in 2003. Notably, this is the first fungal genome that was sequenced. Since then, the ‘1000 Fungal Genome Project’ funded by the Joint Genome Institue (JGI) to sequence 1000 fungal genomes from across the Fungal Tree of Life.³⁶ This project has significantly increased the amount of sequenced fungal genomes.

Recently, the Cox group³⁷ sequenced the genome of thirteen taxonomically well-defined fungi from Hypoxylaceae (Xylariales, Ascomycota) family and one Xylariaceae by using combinations of Illumina and Oxford nanopore technologies or PacBio sequencing. These high quality genome sequences not only satisfy the taxonomic purposes in mycology but also provide opportunities for the study of fungal evolution, host-fungus interactions, as well as the biosynthesis of secondary

metabolites. For instance, more than 750 biosynthesis gene clusters have been found from the thirteen sequenced genomes. These gene clusters include various types of secondary metabolic pathways, such as polyketides, terpenes, peptides, meroterpenoids and alkaloids. And the discovery and characterization of cytochalasan³⁸ and azaphilone³⁹ gene clusters resulted from these genome sequences.

1.3.3 Gene Knockout

Targeted knockout (KO) can result in two outcomes. Either the KO with the designed target, or ectopic intergration occurs elsewhere in the genome (Figure 1.7). Both possibilities result in incorporation of the selectable marker. Often ectopic intergration greatly exceeds targeted incorpation, meaning that tedious screnning is required to find the desired KO transformant.⁴⁰

Figure 1.7 Integration of a knockout cassette.

Fairhead and co-workers⁴¹ developed a split-marker technology, also called the bipartite method, to overcome this problem for S. cerevisiae. This method has also been applied successfully to diverse filamentous fungi.

Split-marker technology requires a mixture of two DNA fragments comprising overlapping sequences of a selectable marker gene. Only by homologous recombination of three crossing-over events can generate a functional marker gene, which allows producing an intact gene targeting cassettes for gene substitution in fungal transformation. These two fragments can be easily obtained from PCR amplification (Figure 1.8).⁴²

Nielsen and co-workers⁴² showed that bipartite knockout results in a higher frequency of correct targeting events compared to that classical transformation of a continuous gene targeting cassettes.

The bipartite method is extremely flexible and can be easily applied in genome manipulations, like promoter replacements and GFP tagging. However, this technology also has a disadvantage.

KO cassette

Target gene

Disrupt target gene Ectopically intergrated

Knockout cassette Transformation

Target gene

KO cassette

Compared with other traditional transformation methods, the split-marker systems (bipartite method) dramatically reduces the frequency of fungal transformation. For example in Magnaporthe grisea, Jeong and co-workers⁴³ showed that the number of transformant obtained from the split marker was smaller than classical methods under the same transformation condition, which is ˃120 and five to 20 respectively. Although the frequency was reduced by about 85 – 96%, they could get two positive transformants with target gene-substitution from bipartite knockout, but none from the traditional approach.

Figure 1.8 Strategy of bipartite knockout, adapted from Nielsen et al., 2006.⁴²

Also in our group, there are successful examples of gene disruption by using the bipartite method, for example in the biosynthetic investigation of cytochalasan H 63 in Magnaporthe grisea (Scheme 1.8).³⁸ Through the gene inactivation of the functional genes (O-methyltransferase, trans-enoyl reductase, O-acetyltransferase, oxidoreductase and P450, respectively), the late-stage biosynthetic pathway of 63 was fully elucidated and reveals that O-methyltyrosine 61 is the true precursor for 63.

Scheme 1.8 The proposed biosynthesis of cytochalasan H 63 by Wang et al., 2019.³⁸

Wild type Target gene

Mar ker

ker gene

Marker gene Mutant

Homologous recombination Split-marker system

1.3.4 Heterologous Expression

Heterologous expression not only can be used to discover the secondary metabolite pathways but also enable the improvement of the yields of natural products. Also, this strategy provides the opportunity in synthetic biology study to produce novel compounds. A good heterologous host features two factors at least: easy cultivation and genetic manipulation. For example, both bacteria (E. coli and Streptomyces sp) and fungi (Aspergillus oryzae, Aspergillus nidulans, and Saccharomyces cerevisiae) are used as heterologous expression platforms.⁴⁴

For instance, the first successful heterologous expression of the polycyclic tetramate macrolactam (PTM) ikarugamycin BGC in E. coli opened one way to investigate cryptic iPKS/NRPS biosynthetic pathways found in other bacteria.⁴⁵ However, it is not possible for bacteria to process eukaryotic introns and bacteria often possess a significant codon bias, therefore bacteria are not the ideal hosts for fungal gene expression.

Yeast (S. cerevisiae) is a suitable host for some fungal BGC expression. Recently, the bostrycoidin (a red aza-anthraquinone pigment) gene cluster was successfully expressed in an engineered S. cerevisiae.⁴⁶ Firstly, the primary metabolism of the S. cerevisiae was optimized for higher flux towards the acetyl- and malonyl-CoA pathways which render a higher concentration of precursor for the bostrycoidin polyketide construction. In addition, a PPTase native to the original producer was cloned into the expression host and co-expressed with the PKS to activate the ACP domain, required for polyketide backbone assembly. Finally, the maximum titer (2.2 mg∙L^‒1) of bostrycoidin production was achieved after 2 days of galactose induction.

The filamentous fungus A. oryzae has been utilized to express fungal gene clusters for many years, and there are many successful examples in our group. For instance, the biosynthetic pathway elucidation of the maleidride byssochlamic acid, the meroterpenoid xenovulene A, the polyketides squalestatin S1 and sorbicillinoids.14,21,47,48 The well-established genetic manipulation approaches in A. oryzae make it a great success as a heterologous expression system to study secondary metabolite biosynthesis and production.⁴⁹

Heterologous expression can also be used to solve tricky biosynthetic problems. For example, heterologous expression in A. oryzae was used to characterise the function of two putative nonheme-iron-dependent enzymes (Mfr1 and Mfr2) involved in the post tailoring steps of squalestatin S1 24 biosynthesis (Scheme 1.9).¹⁴ Co-expression of the oxidase (Mfr1), PKS (Sqhks), hydrolase (Mfm8) and citrate synthase (Mfr3) led to the observation of several oxidised congeners (64 m/z 435 [M‒ H]^‒1, 65 m/z 433 [M‒ H]^‒1, 66 m/z 431 [M‒ H]^‒1 and 67 m/z 447 [M‒

H]^‒1) in LCMS chromatogram. 50 is a substrate for the stepwise oxidations by Mfr1, first make

alcohol 64, then ketone 65, unsaturated ketone 66, and further oxidise to give the epoxide 67.

When additional expressed the oxidase Mfr2 to the system, Liquid Chromatography Mass Spectrometry (LCMS) results showed the oxidised metabolites of 68 m/z 463 [M‒ H]^‒1 and 71 m/z 479 [M‒ H]^‒1. There is a proposed mechanism of acetal formation and epoxide opening during the conversion of 70 to 71.

Subsequently, the copper-dependent oxygenase Mfm1 introduces a hydroxyl group (73) required for later acetylation to give 51 that servers as the substrate for the final acylation reaction catalysed by Mfm4 to afford squalestatin S1 24 (Scheme 1.9).

Scheme 1.9 The proposed oxidative cascade during the biosynthesis of squalestatin S1 24 by Lebe et al., 2019.¹⁴

1.4 Overall Aims

The overall aim of this project is to focus on understanding the biosynthesis of alkyl citrate compounds such as sporothriolide 1 and polyketide polyenes such as trienylfuranol A 2, as well as the adduct sporochartine 3.

Extensive strategies of isotopic labelling, genome and transcriptome sequencing and bioinformatic analysis, gene knockout, heterologous expression and in vitro studies will be performed to delineate the biosynthetic pathway of sporothriolide and trienylfuranol A. In addition, the study will support the proposed biosynthesis of alkyl citrates of piliformic acid and oryzines.

For sporochartine which is proposed to be a consequence of Diels-Alder cycloaddition of sporothriolide and trienylfuranol A, the particular aim is to answer the question of whether it’s an enzymatic (Diels-Alderase) catalysis result or a chemically spontaneous reaction.

2 Biosynthetic Studies of Sporothriolides

2.1 Introduction

Sporothriolide 1 was first isolated and elucidated by Krohn and co-workers⁵⁰ from Sporothrix sp.

(strain No. 700). Lately, the Stadler⁵¹ group published their work about the secondary metabolites from Hypoxylon monticulosum MUCL 54604 (now referred to as Hypomontagnella monticulosa).^37,52 They found that this organism mainly produces sporothriolide 1 together with other analogues (Figure 2.1). Shortly after, the Ouazzani⁵³ group isolated sporothriolide and the structurally more complicated sporochartine from Hypoxylon monticulosum CLL 205 (now referred to as Hypomontagnella spongiphila) isolated from a marine sponge.^37,52 The crystal structure of sporothriolide 1 was first reported by the Ye⁵⁴ group, obtained from the endophyte Nodulisporium sp. A21.

Bioactivity studies have shown that sporothriolide 1 is a potent antifungal agent with EC50 of 11.6

± 0.8 μM against the phytopathogenic fungus Rhizoctonia solani,⁵⁵ while the EC50 of positive control carbendazim was 9.6 ± 0.7 μM.

Figure 2.1 Structures of sporothriolide 1, dihydrosporothriolide 74, sporothric acid 75, deoxysporothric acid 76, isosporothric acid 77 and dihydroisosporothric acid 78.

Sporothriolides probably belong to the alkyl citrate family of metabolites, similar to piliformic acid and the oryzines (Section 1.2.3 ‒ 1.2.5). These compounds are generated from the condensation of fatty acids with a decarboxylated Krebs cycle intermediate, commonly oxaloacetate. Isotopic labelling experiments for deoxysporothric acid 76 using [1-¹³C] and [2-¹³C]

acetate reported by the Ye⁵⁴ group showed that the labelling pattern is consistent with the hypothesis of a fatty acid or polyketide origin. A speculative biosynthetic pathway is shown in Scheme 2.1. First, the C10 fatty acid chain 79 is condensed with C4 oxaloacetate 41, followed by decarboxylation and dehydration to result in intermediate 81, that undergoes lactonization to form deoxysporothric acid 76.

Although sporothriolide 1 has been known as potent antifungal agents for almost two decades, their detailed biosynthesis was still unknown before our work.

Scheme 2.1 Labelling patterns and hypothetical biosynthetic pathways for deoxysporothric acid 76, adapted from Cao et al., 1999.⁵⁴

2.2 Project Aims

Although the labelling patterns for deoxysporothric acid 76 have been reported, the labelling pattern of sporothriolide 1 was unknown. It will be interesting to compare the difference and similarity of these two metabolites. The biosynthetic gene cluster for sporothriolide is also unknown at the start of this project.

Genome sequencing of the producing fungi, supported by transcriptomic studies will be performed to predict the putative biosynthetic gene cluster (BGC) of sporothriolide.

Transcriptomic analysis of H. monticulosa under sporothriolide producing and non-producing conditions will be conducted. Results will be used to identify intron positions and translational start and stop positions. Gene knockout will be used to further confirm the correct BGC for sporothriolide. Heterologous expression will be used to reconstitute the biosynthesis of sporothriolide in A. oryzae. Meanwhile, the function of interesting enzymes will be investigated by expression and in vitro assay.

2.3 Results

2.3.1 Sporothriolide Production from H. monticulosa, H. spongiphila and H.

submonticulosa

H. monticulosa and H. spongiphila are known to produce sporothriolide.^51,53 Here, we cultivated Hypoxylon submonticulosum (now referred to as Hypomontagnella submonticulosa) and found that it was able to produce sporothriolide metabolites as effectively as the other two fungi (Section

4.3.1). H. monticulosa, H. spongiphila, and H. submonticulosa were obtained as gifts from the Stadler,⁵¹ Ouazzani⁵³ and Sumarah⁵⁶ groups, respectively.

2.3.1.1 Producing and Non-Producing Conditions

In order to investigate the expression level of the biosynthetic gene cluster, producing and non- producing fermentation conditions were studied. The cultivation condition of H. monticulosa was originally reported in YMG medium (Table 6.5),⁵¹ the H. spongiphila and H. submonticulosa were previously grown in PDB medium (Table 6.5).^53,56 The organisms were grown in various liquid media in shake culture. At the end of the fermentation, the cultures were extracted into ethyl acetate. The concentrated organic extracts were examined by LCMS. The difference between PDB and YMG fermentation conditions of H. monticulosa was compared in our experiment, but the chemical profile of the secondary metabolites from two cultivation media showed no obvious variation. Finally, we chose ‘PDB medium, 130 rpm, 28 °C, 6 day’ as sporothriolide producing conditions for H. monticulosa and H. spongiphila (Figure 2.2). And the H. submonticulosa producing conditions are ‘PDB medium, 100 rpm, 25 °C, 6 day’ (see LCMS chromatograph in Section 4.3.1).

Figure 2.2 Diode array detector (DAD) chromatograms of H. monticulosa MUCL 54604extracts: A, under producing conditions (PDB medium, 130 rpm, 28 °C, 6 d); and B, under non-producing conditions (DPY medium, 130 rpm,

28 °C, 6 d).

We tested DPY, LB and MMK2 (Table 6.5) media as potential non-producing fermentation conditions. The results showed that no sporothriolide metabolites were produced in MMK2 and DPY media at least up to day 6 of the fermentation. Because more mycelia was harvested in DPY medium, ‘DPY medium, 130 rpm, 28 °C, 6 day’ was used as the non-producing conditions for H.

monticulosa and H. spongiphila (Figure 2.2).

A DAD

B DAD

1

7

DAD (210-600 nm)

A

DAD (210-600 nm)

B

Producing conditions PDB, 130 rpm, 28 °C, 6 d

Non-producing conditions DPY, 130 rpm, 28 °C, 6 d

min 2

1

2.3.1.2 Time Course Study of Sporothriolide Production

In order to quantify the production of compounds, the calibration curve for sporothriolide 1 was made based on a UV integration method. A dilution series of 1 (1.5, 1.0, 0.5, 0.25, 0.125, 0.0625, 0.03175 mg∙mL^‒1) was measured and the corresponding signals for the extracted wavelength at 211 nm (λmax) were integrated (Figure 2.3A). Integrated values were then plotted against the sample concentration and fitted into a straight line (Figure 2.3B). The relationship is linear within the 0.03 – 1.5 mg∙mL^‒1 concentration range. The equation was applied to quantify 1 across different samples.

Figure 2.3 A, the integrated values of different dilutions of sporothriolide 1 solutions at 211 nm from LCMS analysis;

B, calibration curve for sporothriolide 1 quantification.

Timecourse experiments for the three Hypomontagnella wild type strains were conducted to monitor the production of 1. H. monticulosa and H. spongiphila were grown respectively in a flask containing 1 L PDB medium at 28 °C, 130 rpm for 14 days and 10 mL aliquots were taken daily. H. submonticulosa was grown in a flask containing 1 L PDB medium at 25 °C, 100 rpm for 9 days and 10 mL aliquots were taken daily. Aliquots were extracted with equal amounts of ethyl acetate and the organic phase was evaporated under vacuum (Section 6.2.1). Crude extracts were analysed by LCMS (Section 6.2.2) and compound titres were calculated using the previously described equation (Table 2.1). Production of 1 was visualized graphically (Figure 2.4).

The data shows that all three fungi can produce high amounts of sporothriolide 1. The H.

spongiphila and H. submonticulosa strains had the biggest production of 1 at the 6^th day with 181 mg∙L^‒1 and 238 mg∙L^‒1, respectively. However, the highest production amount of H. monticulosa was at the 4^th day with 190 mg∙L^‒1. All strains displayed a dramatic production decrease of sporothriolide at the 7^th day, which indicated a fast degradation and a short life cycle of 1.

Meanwhile, sporothriolide 1 and cometabolites 74 ‒ 78 were purified, characterised by NMR and compared to literature (Chapter 7).

211 nm

Conc. / mg∙mL^‒1 UVint

1.5 132848

1 96828

0.5 46285

0.25 21143

0.125 9113

0.0625 4340

0.03175 1650

UV = 91894C - 942.42 R² = 0.9965 0

50000 100000 150000

0 0.5 1 1.5 2

UV integral

C [mg∙mL^-1]

A B

Table 2.1 Temporal production of sporothriolide 1 from H. monticulosa MUCL 54604, H. spongiphila CLL 205, and H.

submonticulosa DAOMC 242471 cultivated under producing conditions.

Figure 2.4 Kinetic production curve of sporothriolide 1.

2.3.2 The Identification of Multiforisin H

During the investigation of metabolites from H. spongiphila CLL 205, we isolated a known compound multiforisin H 82 (Figure 2.5) which was initially found from Ascomycete Gelasinospora species.⁵⁷ The previous bioactivity study revealed multiforisin H 82 is a potent immunosuppressive agent, with IC50 values of 1.8 ug∙mL^‒1 and 0.9 ug∙mL^‒1 against Con A- (T- cells) and LPS-induced proliferations of mouse splenic lymphocytes, respectively.⁵⁷

H. monticulosa H. spongiphila H. submonticulosa Day Conc. / mg∙L^‒1 Conc. / mg∙L^‒1 Conc. / mg∙L^‒1

1 0 0 0

2 71.3 0 0

3 164.2 56.4 0

4 190.7 162.1 0

5 136.2 98.0 64.5

6 141.5 181.3 238.3

7 32.4 55.6 87.0

8 38.7 39.6 15.6

9 22.2 8.6 0

10 7.7 0 -

11 0 0 -

12 7.4 2.4 -

13 0 0 -

14 0 0 -

0 50 100 150 200 250 300

1d 2d 3d 4d 5d 6d 7d 8d 9d 10d 11d 12d 13d 14d

mg∙L‒1